mcp-server-for-revit
Provides tools for interacting with Autodesk Revit, enabling AI agents to query view information, create and modify elements (walls, rooms, grids, levels, etc.), tag elements, export data, and execute custom C# code within Revit.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-server-for-revitlist all walls in the current project"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-server-for-revit
MCP server (Node.js) to interact with Autodesk Revit from AI assistants (Claude, Cursor, etc.).
This repository contains two things that are important to distinguish:
The "original" MCP (Node bridge) — aligned with the mcp-servers-for-revit ecosystem: exposes Revit operations as MCP tools; the server talks to the official Revit plugin via TCP (default port
8080), not via WebSocket in this bridge flow.The extension of this fork — dual port in the bridge (
REVIT_PLUGIN_PORT/REVIT_SKETCH_PORT) and the separate add-inRevitSketchPoC(C# / WPF) in therevit-plugin/RevitSketchPoC/folder, which listens on another port (default8081) and handles only thecreate_house_from_sketchtool (image interpretation + creation of walls/rooms/doors in Revit), with Ollama or Gemini configurable in the plugin.
[!NOTE] For the full flow, you need the original Revit plugin (port
8080) and, if you use the sketch via MCP or UI, the RevitSketchPoC (port8081). Official plugin instructions: mcp-servers-for-revit. PoC detail:revit-plugin/RevitSketchPoC/README.md.
Security and Git
Do not commit
.envfiles, JSON keys, or local copies with secrets. The repository includes a.gitignorefornode_modules/, .NET artifacts (bin//obj/),.env*, andpluginsettings.local.json.The
deploy/pluginsettings.jsonfile of the PoC must keepGeminiApiKeyempty in Git; use a local copy or follow thepluginsettings.example.jsonas a template.
Setup (MCP client)
Claude Code
claude mcp add mcp-server-for-revit -- npx -y mcp-server-for-revitClaude Desktop — claude_desktop_config.json:
{
"mcpServers": {
"mcp-server-for-revit": {
"command": "npx",
"args": ["-y", "mcp-server-for-revit"],
"env": {
"REVIT_PLUGIN_PORT": "8080",
"REVIT_SKETCH_PORT": "8081"
}
}
}
}Restart the client. The tools icon indicates that the MCP is connected.
Two ports, two plugins
Environment variable | Default value | TCP Destination |
|
| Original plugin — all tools except |
|
| RevitSketchPoC — only |
The create_house_from_sketch tool uses REVIT_SKETCH_PORT (or data.pluginPort in the request). The pluginPort field is not sent to Revit (the bridge removes it). The remaining tools use REVIT_PLUGIN_PORT.
Supported MCP Tools
Tool | Description |
| Active view information |
| Elements in the active view |
| Available family types |
| Selected elements |
| Material quantities |
| Intelligent model query |
| Model statistics |
| Point-based elements |
| Line-based elements |
| Surface-based elements |
| Grid |
| Levels |
| Rooms |
| Dimensions |
| Beam system |
| Delete by ID |
| Operate (select, color, hide, etc.) |
| Color by parameter |
| Tags |
| Export room data |
| Local metadata |
| Send C# to execute in Revit |
| Sketch image → LLM interpretation (Ollama or Gemini in RevitSketchPoC) → walls/rooms/doors |
| Connection test |
Development (Node)
npm install
npm run buildSketch PoC (C# / WPF)
The current and supported implementation is in revit-plugin/RevitSketchPoC/ (UI, MCP TCP router, Ollama/Gemini, Revit transactions). The samples/RevitSketchPoC/ folder may exist as an old reference; follow the plugin README at revit-plugin/RevitSketchPoC/README.md.
Summary flow:
User sends image (ribbon Sketch AI PoC or MCP
create_house_from_sketch).The plugin calls the LLM (Ollama or Gemini according to
pluginsettings.json).Structured response → creation of elements in a
Transactionin Revit.
License
This server cannot be installed
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
MCP server for AI dialogue using various LLM models via AceDataCloud
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to control Unreal E…
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/vascocardosondBIM/MCPServerTest'
If you have feedback or need assistance with the MCP directory API, please join our Discord server